2.7 Nutrient enrichment

High levels of nutrients (N and P) from agricultural runoff (both fertilizers and animal waste), storm water inputs, and treated and untreated sewage have lead to high levels of nutrients and water quality problems in many lakes, streams, and other freshwater habitats (Kiffney et al., 2005). Conversely, in many oligotrophic waters in temperate and arctic latitudes, the addition of organic and inorganic nutrients (N and P) has been used as a method to increase primary and secondary productivity and presumably fish growth and abundance (Benndorf, 1990; Kiffney et al., 2005). Nutrient additions are also widely used in aquaculture in floodplain lakes and ponds to boost fish production.

2.7.1 Nutrient enrichment techniques

A handful of techniques have been developed to enrich nutrient levels and boost productivity of waters including the addition of organic (fish carcasses) and inorganic nutrients (typically N and P). For aquaculture purposes, this often includes diversion of nutrient rich agricultural run off, addition of inorganic nutrients, and addition of manure or other solid organic material (Welcomme and Bartley, 1998). The recent popularity of nutrient enrichment in North America has lead to some new developments for applying these nutrients including "fish blocks", which are produced using fish feed or bone and flesh waste from fish processing plants to time release fertilizers specially designed for stream applications (Sterling et al., 2000; Wipfli et al., 2004; Kiffney et al., 2005). While nutrient enrichment techniques are sometimes termed restoration, they are in essence a form of enhancement that is more often than not used in an attempt to increase the productivity of a water body above historical levels. However, sustaining this level of production is typically dependent on continued addition of nutrients (Guthrie and Peterman, 1988; Gross et al., 1998;).

The addition of nutrients to streams and a lesser extent lakes has become more common place in western North America following a number of studies that demonstrated the importance of anadromous fishes to the productivity of freshwater ecosystems (Krokhin, 1975; Cederholm et al., 1999; Roni et al., 2002; Kiffney et al., 2005). For example, it is estimated that reductions in returning salmon to the Pacific Northwest coastal streams has reduced marine derived N and P delivered by returning adult salmon to six to seven percent of historic levels (Gresh et al., 2000). Studies for other anadromous fishes such as smelt (Osmeridae), shad (Alosa spp.), Atlantic salmon, and sea trout have demonstrated similar results in tributaries to the North Atlantic (Durbin et al., 1979; Garman and Macko, 1998; Lyle and Elliott, 1998).

2.7.2 Effectiveness of nutrient enrichment

Lakes -The effectiveness of these techniques has been examined in numerous studies in both lakes and streams in North America and Europe, though much of the work has occurred in lakes. The experimental manipulation of N and P levels to increase primary and secondary production in lakes has been occurring for many years in Alaska (Kyle, 1994; Mazumder and Edmundson, 2002), midwestern US (Carpenter et al., 1996; Cottingham and Knight, 1995; Vadeboncoeur et al., 2001; Essington and Houser, 2003), western United States (Sarnelle, 1992; Cottingham and Carpenter, 1998; Wurtsbaugh et al., 2001; Budy et al., 1998), British Columbia, Canada, (Hyatt and Stockner, 1985; Stockner and MacIsaac, 1996; Ashley et al., 1997; Johnston et al., 1999; Bradford et al., 2000), Northwest Territories (Welch et al., 1988; Jorgenson et al., 1992), eastern Canada (Clarke et al., 1997; Mazumder et al., 1988), Denmark (Jeppesen et al., 1991), Sweden (i.e. Bjoerk-Ramberg, 1983; Jansson, 1984; Holmgren, 1984; Bjoerk-Ramberg and Anell, 1985; Jansson et al., 2001), Norway (Langeland, 1990), and Japan (Ishida and Mitamura, 1986). Increases in nutrients, phytoplankton (chlorophyll a) and primary production, zooplankton, and benthic macroinvertebrates have been demonstrated in many of these studies.

Fish response to lake enrichment has been more variable and difficult to measure. The vast majority of studies on fish response to lake enrichment have been for sockeye salmon (Oncorhynchus nerka) and most have demonstrated increased growth, juvenile survival, and sometimes adult sockeye returns (Hyatt and Stockner, 1985; Ashley et al., 1997; Kyle, 1994; Stockner and MacIsaac, 1996; Budy et al., 1998; Bradford et al., 2000; Mazumder and Edmundson 2002). A few studies have reported positive results for other species. Johnston et al. (1999) examined the effects of inorganic fertilizers on rainbow trout in a British Columbia lake and found increased growth, reproductive output, but not survival. Essington and Houser (2003) examined yellow perch (Perca flavescens) response to nutrient enrichment and found increased growth and reduced mercury concentrations. Other studies that have focused on examining the effects fish (predator) additions or removals have on food webs and ecosystem components (e.g. Mazumder et al., 1988; Benndorf, 1990; Carpenter et al., 1996; Cottingham and Carpenter, 1998). Studies of nutrient additions of aquaculture ponds and reservoirs in China and Russia have also demonstrated large increases in plankton, benthos, and fish yield following addition of manure or other organic nutrients (Lu, 1992; Welcomme and Bartley, 1998).

Based on studies of fish as well as primary and secondary productivity, the response of a given lake to fertilization depends on a variety of factors including but not limited to: zooplankton and fish species present (Jeppesen et al., 1991; Benndorf, 1990; Carpenter et al., 1996), light levels (Bjoerk-Ramberg, 1983; Benndorf, 1990), stratification (temperature) and depth (Bjoerk-Ramberg, 1983; Jansson, 1984; Wurtsbaugh et al., 2001), whether benthos were sampled (Bjoerk-Ramberg and Anell, 1985; Vadeboncoeur et al., 2001), whether the system was P or N limited (Holmgren, 1984; Jansson et al., 2001), number of years of fertilization (Jorgenson et al., 1992), or top down versus bottom up control of biotic communities (Langeland, 1990; Stockner and Shortreed, 1994).

Streams - The addition of both organic (fish tissue) and inorganic nutrients on streams and their biota does not have as long a history as that for lakes, but there still are several studies that have examined the effects of these two techniques on stream biota (Table 14). Johnston et al. (1990), Ward (1996), and McCubbing and Ward (1997, 2000), in one of the largest and longest term studies on addition of inorganic nutrients to a stream, have detected increases in periphyton, macroinvertebrate abundance, and increases in juvenile coho salmon and steelhead growth and condition. Ashley and Slaney (1997) summarized the results of case studies of inorganic nutrient additions on five different British Columbia watersheds and found increased periphyton, invertebrate biomass, and fish growth following nutrient additions. Wipfli et al., (2003, 2004) in a series of studies in both natural and artificial streams in Southeast Alaska, found increased condition, growth, and production following placement of both salmon carcasses and carcass analogs (artificial carcasses made from fish tissue). Following addition of inorganic phosphorous to an Arctic tundra river, Deegan and Peterson (1992) found an increase in juvenile and adult arctic grayling growth. These studies all suggest positive initial responses to addition of inorganic nutrients to oligotrophic streams.

Addition of organic nutrients in the form of fish carcasses has also been examined in anadromous fish streams in Alaska, British Columbia, Washington and Minnesota. For example, Schuldt and Hershey (1995) found increased P and periphyton following addition of salmon carcasses to a Minnesota stream. In a laboratory experiment, Minakawa et al. (2002) found higher growth of caddis flies on salmon carcasses than on natural leaves. Wipfli et al. (1998; 1999) and Chaloner and Wipfli (2002) found increased chlorophyll a, macroinvertebrates, and growth rate of salmonids following additions of carcasses in artificial and natural stream channels. Studies using stable isotope analysis have demonstrated the importance of salmon carcasses in aquatic food webs for primary productivity, macroinvertebrates, fishes and even riparian vegetation and tree growth (Bilby et al., 1996; Helfield and Naiman, 2001). While additional study is needed, these studies all suggest that the placement of salmon carcasses or an increasing in spawning fish can increase fish growth and possibly survival.

TABLE 14
Summary of published studies on nutrient enrichment in streams

2.7.3 Conclusions - nutrient enrichment

The addition of organic and inorganic nutrients into lakes and ponds has a relatively long history with variable success depending on a number of ecological factors. Nutrient enrichment in streams has a much shorter history, but early studies suggest very positive results. While additional studies are needed particularly for stream nutrient enrichment, we provide the following conclusions and recommendations.

• The addition of inorganic and organic nutrients to oligotrophic streams and lakes can lead to increased growth and production of algae and zooplankton and in some cases increased fish growth.

• Continued addition of nutrients or an increase in natural nutrient delivery (recovery of depressed anadromous fish runs) is needed to maintain elevated production.

• Many factors affect the success of projects and need to be evaluated prior to nutrient enrichment including: base line nutrient status (what nutrients are limiting productivity?), species composition of plankton, macroinvertebrates and fishes (top down versus bottom up control), physical limnology of lake (e.g. depth, stratification, temperature, light).

Additional discussion of determining whether nutrient enrichment is appropriate and methods for monitoring and evaluating responses are provided in Ashley and Slaney (1997) and Kiffney et al. (2005).

2.8 Miscellaneous habitat rehabilitation methods

There are several less common methods of habitat rehabilitation including: beaver reintroduction or removal, streambank debrushing, bank protection, and habitat protection. We discuss these here as they do not easily fit into the categories we have described and because may or may not be habitat rehabilitation, depending on the project goals.

2.8.1 Beaver reintroduction or removal

Beaver dams alter the hydrology and geomorphology of stream systems and affect habitat for fishes as well as fish diversity (Snodgrass and Meffe, 1998; Pollock et al., 2003). They also influence the rates of groundwater recharge and stream discharge, retain enough sediment to cause measurable changes in valley floor morphology, and enhance stream habitat quality for many fishes (Pollock et al., 2003). Historically, beaver dams were frequent in small streams throughout most of the Northern Hemisphere. The cumulative loss of millions of beaver dams has affected the hydrology and sediment dynamic of stream systems. The cumulative hydrologic and geomorphic affects of the loss of millions of wood structures (beaver dams) from small and medium-sized streams are not entirely known. This is particularly important in semiarid climates, where the elimination of beaver and beaver dams has likely exacerbated effects of other land use changes, such as livestock grazing, and possibly accelerated incision and the subsequent lowering of groundwater levels and ephemeralization of streams (Pollock et al., 2003).

Beaver reintroduction has been proposed as a method of restoring the ecological functions described above and has in fact been attempted on a limited basis in the Europe, Russia, Mongolia, and North America. Beaver reintroductions in the USA suggest that rapid recolonization, dam construction, and changes in physical habitat occur following reintroduction, assuming the animals are not harvested or consumed by predators (Apple, 1985; Albert and Trimble, 2000; McKinstry et al., 2001; McKinstry and Anderson, 2002). For example, McKinstry and Anderson (2002) reported successful reintroduction of beavers in 13 of 14 sites. They indicated that best success was gained with beavers greater than 2 years old, as mortality rates among younger beaver were extremely high. Merely reducing or banning of commercial or recreational harvest (trapping) of beaver has lead to them slowly recolonizing many areas in the United States. Studies in Poland have also shown that beavers rapidly colonize habitats following beaver reintroduction (Zurowski and Kaspercyzk, 1988).

While beavers, beaver dams, and associated ponds are known to provide a number of benefits to riparian areas, floodplains, and aquatic ecosystems and particular benefits for juvenile Pacific salmon (Pollock et al., 2004), some fisheries management programmes have recommended removal of beavers and beaver dams to enhance resident trout populations. Unpublished studies in the mid-twentieth century suggested this as an effective technique for improving brook trout habitat in Wisconsin (USA) streams (Avery, 2004). In a review of 103 trout habitat improvement projects conducted in Wisconsin between 1953 and 2000, Avery (2004) reported that beaver dam removal resulted in the highest success rate and largest increase in brook and brown trout numbers of any project type they examined. Avery did not examine nonsalmonid fishes or other ecological factors and his results suggest that beaver dam removal may be effective for a narrow set of objectives such as increasing resident brook trout numbers. However, if the goals are ecosystem recovery or restoration, then reintroduction or protection of beavers appears to be more successful strategy than removal and would benefit many native fishes and other biota.

2.8.2 Bank protection (riprap and bioengineering)

Riprap (large rock placed to protect the bank), riparian plantings and other bank protection methods are often incorporated as part of habitat rehabilitation efforts though their objectives are not always to improve fish habitat. Riparian rehabilitation activities that utilize living materials such as trees and shrubs for bank protection and other forms of fluvial geomorphic control are termed bioengineering (Schiechtl and Stern, 1997). Often the purpose of such activities is to rapidly increase both aboveground and belowground biomass such that the erosion of the underlying substrate is minimized during floods. Bioengineering frequently involves the use of both living and nonliving materials to create a desired feature. Thus gabions, groins, sills and other bank protection structures are planted with living material such as live cutting of willow to restore certain riparian functions. The numerous bioengineering techniques that can be applied towards bank protection are described in detail elsewhere (Gray and Sotir, 1996; Schiechtl and Stern, 1996, 1997). It is important to note that bank protection is a mitigation strategy to protect infrastructure or areas of human uses. It is clearly effective strategies for these objectives and may prevent erosion, but improving habitat is typically a side benefit of these techniques.

The effectiveness of bank hardening, planting, and bioengineering techniques at stabilizing banks has been relatively well documented (RSPB et al., 1994; Schmetterling et al., 2001). In some highly degraded systems, bank protection may prevent further erosion and provides benefits for fishes (O’Grady et al., 2002; Schmetterling et al., 2001). However, it does not provide habitat for multiple life stages and can disrupt natural channel processes such as channel migration and development of natural riparian vegetation (Schmetterling et al., 2001). Moreover, higher diversity of fishes in natural (woody debris, sand and grass banks) versus artificial habitats (riprap banks, bridge piling, groins) has been documented (Madejczyk et al., 1998; Peters et al., 1998; Schmetterling et al., 2001). Those bank protection techniques that include bioengineering appear to be more beneficial to fisheries resources than hard structures. For example, Peters et al. (1998) found that bank protection with woody debris had higher fish densities than those that use only rock, but still had lower fish densities and woody debris levels than natural stream banks. These studies suggest that when necessary, bank protection that incorporates woody debris, willow plantings or similar items (bioengineering) appear be more beneficial to fisheries resources than hard structures, though they may not achieve the benefits of more natural stream banks. In our view bank protection is not really a form of habitat improvement and should be used cautiously in habitat rehabilitation activities.

2.8.3 Brush removal

Debrushing a stream or bank is another technique that seems to be in direct contrast to many other efforts focused on restoring riparian conditions and streambank rehabilitation. A major assumption is that removal of dense brush that shades the stream will lead to an increase in macroinvertebrates and thus fish production, which is in part supported by results of previous studies on logging (see Hicks et al., 1991 for a review). Brush removal and trimming of vegetation are also common methods for manipulating habitats for birds, wildlife, and sometimes to control localized flooding (RSPB et al., 1994). Brush removal is also sometimes used as a method for controlling invasive exotic riparian species and allow for recolonization of native vegetation (see section on riparian rehabilitation). The cutting of weeds to form a sinuous channel while leaving bank brush and vegetation is a technique for fisheries rehabilitation common in Denmark and other countries (Iversen et al., 1993).

The effects of brush removal on fisheries resources have been inconsistent. O’Grady (1995) found higher level of Atlantic salmon age 1+ and brown trout age 1+ smolts following brush removal in River Arrigle in southeast Ireland. In contrast, Avery (2004) found that brush removal alone or combined with placement of cover logs provided little benefit to resident trout populations in Wisconsin streams. Some of the disparate results of these studies likely have to do with changes in light, organic inputs, and stream temperatures following removal of brush. These factors, unfortunately, were not examined in the studies we located on brush removal. Given the disparate results of these studies, drawing firm conclusions on the effectiveness of this technique is difficult; therefore, brush removal should be used with caution as habitat rehabilitation technique.

2.8.4 Habitat protection

While not typically considered a form of habitat rehabilitation, habitat protection is often the most cost-effective conservation method, as it is typically cheaper to protect habitat than to try and restore it later (NRC, 1992; Roni et al., 2002). This concept is particularly important in undisturbed areas that are threatened by development or human uses. Habitat protection is often considered the first step when prioritizing habitat rehabilitation activities (Dominquez and Cederholm, 2000; Roni et al., 2002). This often occurs in the form of land acquisitions (purchasing land), conservation easements (buying rights to development but not the land itself), or passing laws to protect areas from further development (Lucchetti et al., 2005).

Although, the effectiveness of habitat protection strategies is generally apparent, monitoring and evaluation of their effectiveness is rarely conducted or published. Lucchetti et al. (2005) in a review of effectiveness of acquisitions and conservation easements in the United States, found no published evaluations of effectiveness. However, Lucchetti et al. (2005) provide guidelines for determining the level of monitoring and evaluation needed for various types of land acquisitions and conservation easements. As with other techniques, the lack of specific information on effectiveness does not mean they are ineffective. Clearly, the protection of high quality and important habitats is important throughout the world, and this underscores the need for thorough planning, careful implementation, and rigorous monitoring of these activities.

2.8.5 Conclusions - miscellaneous rehabilitation techniques

• Beaver reintroduction is an effective strategy for restoring floodplain and slow water habitats.

• Beaver removal may be effective for improving habitats for individual species, but can be detrimental to reestablishing natural processes and maintaining species and habitat diversity.

• Bank stabilization techniques can successfully reduce erosion, inhibit natural channel migration, and can be detrimental to fisheries resources.

• Similar to beaver removal, bank debrushing can improve conditions for specific fishes, but in general is detrimental to many fisheries resources and reestablishing natural processes.

• While not typically considered a rehabilitation technique, habitat protection should be considered prior to rehabilitation efforts.

• Few studies have evaluated habitat protection, but potential benefits are clear and it may be the most effective strategy in many areas.